arXiv subjects

Rohan Singh

Publications and source records attributed to Rohan Singh.

15 recordsLinked to original sources

arXiv2026

Conditional Reliability of Toxicity Signals for Multilingual and Code-Mixed Abuse Detection

Moderation systems increasingly rely on external toxicity tools, but those tools are unreliable under code-mixing, transliteration, slang, and language mismatch. We study the \emph{conditional reliability} of toxicity priors in Indian multilingual and code-mixed short text: English toxicity, Indic abuse, and rule-based severity cues can be useful evidence, but only in some linguistic and abuse-severity contexts. We propose ToxGate, a trust-fusion head that conditions each auxiliary signal on the encoder representation before adding it to the prediction state. Across three short-text abuse datasets, four transformer encoders, and five seeds per setting, ToxGate improves over matched plain encoders in 10 of 12 in-domain settings and 7 of 8 transfer settings. The largest and most interpretable gains occur in high-risk moderation slices, including explicit slurs, violent threats, and cross-dataset transfer. The broader lesson is that moderation systems should treat external toxicity tools and priors as conditional evidence rather than fixed features or ground truth, in focused ablations, source-specific gating gives the strongest results in transfer, severe-abuse slices, and high-risk triage.

cs.CL↗

arXiv2026

Operationalizing Cybersecurity Governance for Mitigation Planning with Attack-Path Modeling and Reinforcement Learning

We address a fundamental challenge in cybersecurity operations of translating governance frameworks into actionable mitigation decisions under realistic resource constraints. Frameworks such as the NIST Cybersecurity Framework (CSF) provide widely adopted measures of organizational maturity, but do not directly support the selection and prioritization of defensive strategies against adversarial behavior. We present a system that operationalizes governance frameworks by mapping CSF maturity assessments into MITRE ATT\&CK mitigation capabilities, which enables direct integration of organizational security posture with adversary-informed defensive planning. To manage adversary complexity, we employ a Variable-Order Markov Model (VOMM) trained on observed ATT\&CK technique sequences to enable scalable adversary simulation within a Deep Reinforcement Learning (DRL) environment. We reconstruct likely attack paths and defensive responses using beam search, and then jointly optimize mitigation selection under explicit budget constraints. Our environment supports concurrent adversaries and realistic mitigation costs. Across multiple reward formulations and configurations, we show that the approach produces stable policies, meaningful cost-risk trade-offs, and interpretable mitigation plans aligned with organizational maturity. These results demonstrate that adversary-aware DRL can generate practical, resource-constrained defense strategies grounded in real-world frameworks and threat behavior.

cs.CR↗

arXiv2026

Fauna Sprout: A lightweight, approachable, developer-ready humanoid robot

Recent advances in learned control, large-scale simulation, and generative models have accelerated progress toward general-purpose robotic controllers, yet the field still lacks platforms suitable for safe, expressive, long-term deployment in human environments. Most existing humanoids are either closed industrial systems or academic prototypes that are difficult to deploy and operate around people, limiting progress in robotics. We introduce Sprout, a developer platform designed to address these limitations through an emphasis on safety, expressivity, and developer accessibility. Sprout adopts a lightweight form factor with compliant control, limited joint torques, and soft exteriors to support safe operation in shared human spaces. The platform integrates whole-body control, manipulation with integrated grippers, and virtual-reality-based teleoperation within a unified hardware-software stack. An expressive head further enables social interaction -- a domain that remains underexplored on most utilitarian humanoids. By lowering physical and technical barriers to deployment, Sprout expands access to capable humanoid platforms and provides a practical basis for developing embodied intelligence in real human environments.

cs.RO↗

arXiv2026

Spectral Tailoring of Inhomogeneous Optical Response Using Two-Dimensional Coherent Spectroscopy

Controlling the coherent optical response of inhomogeneous ensembles is a key challenge in advancing light-matter interaction engineering. We present a comparative study of two spectral tailoring approaches using two-dimensional coherent spectroscopy (2DCS): the prepulse and double-pulse (DP) methods. In the prepulse scheme, a high-intensity pulse induces Rabi oscillations, modulating the 2D spectral amplitude and lineshape when its spectral bandwidth matches the ensemble full width at half maximum (FWHM). To overcome this limitation, the DP method employs variable inter-pulse delay to generate predetermined periodic spectral modulation without bandwidth constraints. Moreover, tuning the relative phase between DP pulses allows selective switching of frequency components, enabling controlled enhancement or suppression of distinct spectral features. These observations highlight that, while the prepulse approach is constrained by spectral bandwidth, the DP method provides a more versatile and reliable route to manipulate the coherent optical response of inhomogeneous ensembles. We are hoping these findings might stimulate further research in optical switching and coherent storage for quantum memory devices using inhomogeneous ensembles.

physics.optics↗

arXiv2026

An integration-free method for calculating curl and divergence in space plasmas using multi-spacecraft data

The knowledge of local spatial gradients (curl, divergence etc.) is crucial to examine the three-dimensional variation of flow fields including velocity and magnetic fields in space plasmas like the solar wind. Here we propose a simple method to calculate the same using the in-situ data of multi-spacecraft systems. Unlike the popular Curlometer method which depends on the vector integration theorems, our integration-free method is based on the construction of a local orthonormal coordinate system and the associated finite difference approximations. The Curlometer is applicable to a four spacecraft system arranged in a tetrahedron and yields a single volume-averaged estimate of the curl. Using our proposed method over 107 intervals of MMS (NASA) data, on the other hand, we successfully calculate the spatial derivatives at the position of each spacecraft of the tetrahedron and a three-spacecraft (non-collinear) subset of the same. The average value of all the curls calculated for a given tetrahedron shows an excellent agreement (correlation coeffcient ~ 0:99) with the curls calculated using Curlometer formula. The quality of the calculated curl (using our method) is found to improve if the spacecraft configuration approaches a regular tetrahedron. The current framework facilitates investigation of turbulent heating rates and the exploration of local flow features like Beltramization, existence of current sheets, etc., in present and future multi-spacecraft mission, including those involving more than four spacecraft.

physics.space-ph↗

arXiv2025

Excitation-pulse intensity mediated control of coherent nonlinear optical response of a V-type system

V-type three-level systems, where two excited states share a common ground state, serve as fundamental models for exploring coherent light-matter interactions in a range of quantum systems, from atomic gases to semiconductor nanostructures. In this work, we investigate the coherent evolution of such a system under strong femtosecond-pulse excitation by numerically solving the optical Bloch equations. Our analysis shows that the coherent evolution of a three-level system critically depends on the product of the excitation-pulse duration and energy separation between the excited states. Building on this understanding, we extend our analysis to simulate two-dimensional coherent spectra in a high-intensity regime. We demonstrate a control over the coherent pathway contributions to the nonlinear optical response of a V-type system by varying the intensity of the excitation pulses. This control is manifested through the ability to selectively turn individual spectral features on or off in the 2D spectra, each corresponding to distinct quantum pathways. Furthermore, the pulse intensities are varied to precisely adjust the phase of these peaks. Our approach provides a simple and robust framework for achieving control of coherent response of multilevel systems.

quant-ph↗

arXiv2025

Quantitative Lineshape Analysis for Arbitrary Inhomogeneity in Two-Dimensional Coherent Spectroscopy

Two-dimensional coherent spectroscopy (2DCS) provides simultaneous measurement of homogeneous and inhomogeneous linewidths through quantitative lineshape analysis. However, conventional lineshape analysis methods assume Gaussian inhomogeneity, limiting its applicability to systems with non-Gaussian inhomogeneity. We present a quantitative lineshape analysis method incorporating arbitrary inhomogeneity using a bivariate spectral distribution function in 2DCS simulations. An algorithm is developed to extract the homogeneous linewidth and arbitrary inhomogeneous distribution from experimentally-measured 2D spectrum. We demonstrate this framework for a quantum-well-exciton resonance with non-Gaussian inhomogeneity. This work broadens the scope of quantitative lineshape analysis for studying materials with non-Gaussian inhomogeneity.

physics.optics↗

arXiv2025

Indirect Excitons and Many-body Interactions in InGaAs Double Quantum Wells

Spatially indirect excitons in semiconductor quantum wells are relevant to basic research and device applications because they exhibit enhanced tunability, delocalized wave functions, and potentially longer lifetimes relative to direct excitons. Here we investigate the properties of indirect excitons and their coupling interactions with direct excitons in asymmetric InGaAs double quantum wells using optical multidimensional coherent spectroscopy and photoluminescence excitation spectroscopy. Analyses of the spectra confirm a strong influence of many-body effects, and reveal that excited-state zero-quantum coherences between direct and indirect excitons in the quantum wells dephase faster than the much higher-energy single-quantum coherences between excitonic excited states and ground states. The results also suggest an important energy-dependent role of continuum states in mediating system dynamics, and they indicate that dephasing mechanisms are associated with uncorrelated or anticorrelated energy-level fluctuations.

cond-mat.mes-hall↗

arXiv2024

HITgram: A Platform for Experimenting with n-gram Language Models

Large language models (LLMs) are powerful but resource intensive, limiting accessibility. HITgram addresses this gap by offering a lightweight platform for n-gram model experimentation, ideal for resource-constrained environments. It supports unigrams to 4-grams and incorporates features like context sensitive weighting, Laplace smoothing, and dynamic corpus management to e-hance prediction accuracy, even for unseen word sequences. Experiments demonstrate HITgram's efficiency, achieving 50,000 tokens/second and generating 2-grams from a 320MB corpus in 62 seconds. HITgram scales efficiently, constructing 4-grams from a 1GB file in under 298 seconds on an 8 GB RAM system. Planned enhancements include multilingual support, advanced smoothing, parallel processing, and model saving, further broadening its utility.

cs.CL↗

arXiv2024

Coherent Nonlinear Optical Response for High-Intensity Excitation

Calculation of the coherent nonlinear response of a system is essential to correctly interpret results from advanced techniques such as two-dimensional coherent spectroscopy (2DCS). Usually, even for the simplest systems, such calculations are either performed for low-intensity excitations where perturbative methods are valid and/or by assuming a simplified pulse envelope, such as a $\delta$-function in time. Here, we use the phase-cycling method for exact calculation of the nonlinear response without making the aforementioned approximations even for high-intensity excitation. We compare the simulation results to several experimental observations to prove the validity of these calculations. The saturation of the photon-echo signal from excitons in a semiconductor quantum well sample is measured. The excitation-intensity dependent measurement shows nonlinear contributions up to twelfth order. Intensity-dependent simulations reproduce this effect without explicitly considering higher-order interactions. Additionally, we present simulation results that replicate previously-reported experiments with high-intensity excitation of semiconductor quantum dots. By accurately reproducing a variety of phenomena such as higher-order contributions, switching of coherent signal, and changes in photon-echo transients, we prove the efficacy of the phase-cycling method to calculate the coherent nonlinear signal for high-intensity excitation. This method would be particularly useful for systems with multiple, well-separated peaks and/or large inhomogeneity.

physics.optics↗

arXiv2023

Exciton-exciton Interactions -- A Quantitative Comparison Between Complimentary Phenomenological Models

Many-body interactions (MBIs) such as exciton-exciton interactions significantly affect the optical response of semiconductor nanostructures. These interactions can be rigorously modeled through microscopic calculations. However, these calculations can be computationally intensive and often lack physical insights. An alternative is to use phenomenological many-body interaction models such as the modified optical Bloch equations (MOBEs) and the anharmonic oscillator (AO) model. While both these models have separately been used to interpret experimental data, to the best of our knowledge, an explicit and direct correspondence between these models has not been established. Here, we show the empirical equivalence between these two complimentary MBI models through two-dimensional coherent spectroscopy simulations. A quantitative correspondence between the MBI parameters used in the two models are obtained. We also perform a quantitative comparison of these MBI models with experiments, which highlights the usefulness of these phenomenological models in interpreting experimental results.

cond-mat.mes-hall↗

arXiv2018

Transient Spectroscopy of Glass-Embedded Perovskite Quantum Dots: Novel Structures in an Old Wrapping

Semiconductor doped glasses had been used by the research and engineering communities as color filters or saturable absorbers well before it was realized that their optical properties were defined by tiny specs of semiconductor matter known presently as quantum dots (QDs). Nowadays, the preferred type of QD samples are colloidal particles; however, there is still a number of applications that would benefit from the availability of high-quality glass-based QD samples. These applications include fiber optics, optically pumped lasers and amplifiers, and luminescent solar concentrators. Here we conduct optical studies of a new type of all-inorganic CsPbBr3 perovskite QDs fabricated directly in glasses by high-temperature precipitation. These samples are scattering free and exhibit excellent waveguiding properties. However, the presently existing problem is their low room-temperature emission quantum yields (QY) of only 1-2%. Here we investigate the reasons underlying the limited QY by conducting transient photoluminescence (PL) and absorption measurements across a range of temperatures from 20-300 K. We observe that the low-temperature PL QY of these samples can be as high as ~25%. However, it quickly drops with increasing temperature. Interestingly, experimental observations cannot be explained in terms of a thermally activated nonradiative rate but rather suggest the existence of two QD sub-ensembles of emissive and completely nonemissive particles. The temperature-induced variation in the PL efficiency is likely due to a structural transformation of the QD surfaces or interior leading to formation of trapping sites or nonemissive phases resulting in conversion of emissive QDs into nonemissive. Thus, future efforts on improving emissivity of glass-based perovskite QD samples might focus on approaches for extending the range of stability of the low-temperature up to room temperature.

cond-mat.mes-hall↗

arXiv2018

Dephasing of InAs quantum dot p-shell excitons using two-dimensional coherent spectroscopy

The dephasing mechanisms of p-shell and s-shell excitons in an InAs self-assembled quantum dot ensemble are examined using two-dimensional coherent spectroscopy (2DCS). 2DCS provides a comprehensive picture of how the energy level structure of dots affects the exciton dephasing rates and recombination lifetimes. We find that at low temperatures, dephasing of s-shell excitons is lifetime limited, whereas p-shell excitons exhibit significant pure dephasing due to scattering between degenerate spin states. At elevated temperatures, quadratic exciton-phonon coupling plays an important role in both s-shell and p-shell exciton dephasing. We show that multiple p-shell states are also responsible for stronger phonon dephasing for these transitions

cond-mat.mes-hall↗

arXiv2014

Coherent Excitonic Coupling in an Asymmetric Double InGaAs Quantum Well Arises from Many-Body Effects

We study an asymmetric double InGaAs quantum well using optical two-dimensional coherent spectroscopy. The collection of zero-quantum, one-quantum, and two-quantum two-dimensional spectra provides a unique and comprehensive picture of the double well coherent optical response. Coherent and incoherent contributions to the coupling between the two quantum well excitons are clearly separated. An excellent agreement with density matrix calculations reveals that coherent interwell coupling originates from many-body interactions.

cond-mat.mes-hall↗

arXiv2012

Correlation and dephasing effects on the non-radiative coherence between bright excitons in an InAs QD ensemble measured with 2D spectroscopy

Exchange-mediated fine-structure splitting of bright excitons in an ensemble of InAs quantum dots is studied using optical two-dimensional Fourier-transform spectroscopy. By monitoring the non-radiative coherence between the bright states, we find that the fine-structure splitting decreases with increasing exciton emission energy at a rate of 0.1 $μ$eV/meV. Dephasing rates are compared to population decay rates to reveal that pure dephasing causes the exciton optical coherences to decay faster than the radiative limit at low temperature, independent of excitation density. Fluctuations of the bright state transition energies are nearly perfectly-correlated, protecting the non-radiative coherence from interband dephasing mechanisms.

cond-mat.mes-hall↗